codebase-overviewlisted

Generate a visual overview of a codebase as a Mermaid flowchart. The objective is to spot, for each entry point, the **landing point** where information is ultimately processed. Only two things matter: where information *enters* the codebase (the entry point) and where it *ends up being processed* (the landing point — either custom code or a third-party library). Everything in between is plumbing and must be discarded, because tracing every package-to-package hop produces an unreadable diagram. If a specific entry point or package was provided as an argument, restrict the analysis to it. Otherwise, include all detected entry points. Requested scope (optional): $ARGUMENTS ## Scale guard Count detected entry points before tracing. If there are more than **30 entry points** and no scope argument was given: - Group entry points by package/module prefix and produce one diagram per group, or - Ask the user to narrow the scope with a specific entry point or package argument. Never emit a diagram with more than 40 nodes total — it becomes unreadable regardless of grouping. ## The three node roles (language-agnostic) Every node in the diagram is exactly one of three roles. No other nodes exist. - **Entry point** — where information enters the codebase: `main()` functions, HTTP route definitions, CLI command handlers, message/queue consumers, exported public API functions. - **Landing point — custom code** — the terminus of the flow inside the project: the **deepest internal p